IRB 2.0

The new regulatory regime that we call Basel IV, will be introducing some updates to the old internal ratings-based approach (IRB) framework. Clearly, several new requirements introduced into this new framework including the output threshold, which would need to be measured against the risk-weighted assets (RWAs) by the New Standardized Approach for the sake of setting up the threshold maximum at 72,5% as for the measured total RWAs.

However, most non-IRB banks in the past had been exposed to higher standards of entry to the IRB approach. Moreover, for those attempting to become an IRB institute, there was this conscious fear of being dependent on the models. Nevertheless, these barriers have been somewhat lowered by the recent IFRS9 implementation as these data and other model governance requirements were less strict and was stipulating a much shorter implementation phase – in case of some banks it was even less than a year. In comparison, the IRB implementations used to be much longer extending up to the three years.

We do expect that the number of banks with the interest to become IRB institutes in all SSM territories will increase significantly in the coming years. There are multiple reasons for this expectation, which we would like to elaborate in this article. We will also elaborate on the challenges and benefits that the banks might experience due to the changes within the IRB framework (Basel III: Finalising the post-crisis reforms (BCBS 424)) that would kick in by 2022.

It would be for many retail banks difficult to stand up against the competition if they cannot increase their capital levels to grow their retail (retail mortgages and retail SME) portfolios. In addition, given the ending of the Quantitative Easing by the FED and by the European Central Bank (ECB) expectedly in the near future, it will be clear to all parties that in the new financial order capital would be a rare commodity. Consequently, the long but more realistic method to grow the portfolios for these retail loans will be to achieve the IRB status and make use of the more benevolent risk weights for the estimation of the capital requirements. There are other benefits of the IRB methods – the improved risk sensitivity to the risk levels of the credit portfolios and the institute-wide incentivized better risk practices. However, implementing IRB will come with many challenges, and the institution should be planning to achieve this goal in the long term. The key challenges we can group in three categories; data requirements, the required knowledge, and the management of regulatory and bank-intern expectations.

All of the challenges have some impact on the business and the institutes’ infrastructure so that the business decision makers should be aware of them.

Data requirements

This could be the most difficult hurdle among others. Independent of the size of the bank’s portfolio, the bank needs to calculate the through the cycle estimates for the Probability of Default (‘PD’) and the Downturn Loss Given Default (‘LGD’) models under the IRB methods. The current CRR rules stipulate the use of the lengthy historical observation periods; for the retail mortgages, you would be looking back for at least five years average default rates for the PD models and for at least seven years of the defaulted loss data for the LGD models. However, for the upcoming IRB rule changes, which we title with the general acronym of IRB 2.0, the most recent European Banking Authority (EBA) papers suggest that we would need to be using as a minimum a data collection/observation period not any less than 20 years for the Downturn LGD estimation. Though, this is a novel idea for most of the retail mortgage portfolios of the European banks unless they make use of the external data providers, this might be not attainable target as either some fifteen years ago such portfolios were too small, or the loans had different product features such that the representativeness could not be established. Consequently, there are such technical hurdles in place. However, in the last fifteen years the data bureaus or other data providers have expanded their activities gradually and in fact can substitute for many such portfolios the lack of the sufficient lengthy loss data periods. Luckily, the new regulatory guidance suggested by the EBA clarified the way that the external data are going to be used for the development of the IRB models.

Required knowledge

Obviously, IRB implementation requires significant knowledge and experience. However, contrary to the times when we were first time implementing the IRB models, we are equipped with a much more thorough guidance compared what was available in 2005/2006 when several leading banks have been preparing for their IRB application. Per say, we did not have access to all the surveys, consultancy papers, and all other technical material that had been produced in the last two to three years by the EBA and ECB. ECB was not ruling over the SSM in 2005/2006. Thru the recent engagements such as ‘Targeted Review of Models’ ECB had signaled the simple fact that they have one clear objective in mind, to make all the IRB models independent in which territory they are deployed, comparable in terms of their technical method prowess to avoid inconsistencies and use of unreliable models for the capital charge estimation. Therefore, today we do have access to much more detailed technical guidance then what we used to have in the early days of the IRB implementation.

However, this is not the key change in the rule set at the times of the IRB 2.0. The key difference is that we do not need to have full coverage in terms of the IRB model coverage for all asset classes; i.e. in Germany, the coverage had a target ratio of 92%, so meaning that out all of the credit exposures at least 92% should be subordinated to the IRB rules under the current CRR. This is history under IRB 2.0, which would be replacing the current IRB rules by 2022. We have a brave new game changer, which gives us the chance to have IRB compliance at portfolio level. Consequently, any bank can have a number of IRB portfolios where other portfolios can still be permanently using the Standardized Approach, i.e. only 40% of a bank’s portfolio could be under the IRB approach and the rest under the Standardized. Certainly, Banks would not need to hire an army of experts to develop models and validate them, given the level of knowledge most banks have acquired thanks to IFRS9 implementation. There will be some relief to manage the costs somewhat in a more reasonable fashion. Most managers have now a clear idea about the PD and LGD models, which were introduced even in the simplest banks under the Standardized Approach due to the implemented IFRS9 impairment frameworks for the calculation of the IFRS9 governed provision levels. Therefore, there must be much less resistance to take the next step to move on with the IRB 2.0 for the capital charge calculation.

In contrast to the early years of the IRB implementations, we do have now in almost all SSM countries a much larger number of technical experts who can with ease handle the model development and validation requirements. Not to mention that most banks have gained significant experience thru IFRS9 impairment framework implementation to conduct such comprehensive projects influencing almost all bank activities. So acquiring the knowledgeable resource base is as well much less problematic.

Management of the Regulatory expectations

With the arrival of the supervision of the ECB, the IRB bank universe had to undergo a change of orientation as they quite rapidly realize that the ECB model review requires a more complete adherence to the CRR standards than what they might have eyed in the past. ECB has certainly improved dramatically the standards of the existing IRB models, and going forward they would be in the steering wheel to control the process of the IRB applications. The Joint Supervisory Teams would be as well relying on the local regulatory experts however, there would be no arbitrage allowed across SSM territories. So having a clear and concise manner of communicating the goals of the IRB programme to the ECB by structuring a reasonable realistic implementation plan would be the best that a bank’s IRB project manager can expect to achieve. Such an application normally takes a minimum of three years, and the regulatory challenges might cause delays. However, these belong to the rule set of achieving and keeping the IRB status, meaning banks have to learn to deal with the ECB expectations. In addition, for this end, we do have ample documents, letters and all sorts of communiques that would support banks establishing a solid communication policy with regulators.

The bigger picture

Calling the IRB project a capital saving programme would be an obvious misnomer as the benefits are much deeper and far-reaching. Achieving IRB status would provide benefits such as better risk governance, diligent planning and portfolio management perspectives, harvesting model and system synergies and improved data use – and not just capital reduction.

Banks especially active in the retail segment would be much less competitive if they continue to measure their capital requirements using the Standardized Approach. As the difference would be that by switching to the IRB status, you can grow your portfolio size up to double with the same level of capital whereas your conversion to the IRB would not cost you significant investments. So eventually in the retail segment specifically, we do have the expectation that institutes would in the future only sustain their market share if they can convert to the IRB approach as the return on capital under the IRB regime provides them superiority from the regulatory perspective. This is a trend, which has already started in many SSM countries and would likely be an on-going trend in the next years.

Please contact us for more information:

Kaan Aksel

Telefon: +49 69 9585 5874

Mobil: +49 160 956 48160

kaan.aksel@pwc.com

 

Hinterlassen Sie einen Kommentar

Ihre E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert.